fast and energy-efficient computation of HOG features
C++ Makefile
Switch branches/tags
Nothing to show
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Failed to load latest commit information.
images_640x480
include
src
tests
Makefile
README.md

README.md

libHOG

initial configuration: 0. install OpenCV 0. point Makefile to OpenCV 0. mkdir tests_output; mkdir build; 0. make

basic example on how to use libHOG in C++:

//computing HOG features on my_image.jpg
cv::Mat img = cv::imread("my_image.jpg");
libHOG lHOG();
lHOG.compute_pyramid(img);
//HOG pyramid is located here: lHOG.hogBuffer_blocks[scale]
//pyramid dimensions: lHOG.hogHeight[scale], lHOG.hogWidth[scale]
//number of scales: lHOG.hogHeight.size()

see src/libHOG_cli.cpp for an additional example of how to use libHOG

linking your project with libHOG: include libHOG.h, and dynamically link with libHOG.so

planned functionality: 0. Matlab API (similar to the interface in voc-release5 featpyramid.m)